Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!cs.utexas.edu!uunet!comcon!roy
From: roy@comcon.UUCP (Roy M. Silvernail)
Newsgroups: comp.binaries.ibm.pc.d
Subject: Re: Filters (was ABE Source & introduction)
Message-ID: <344@comcon.UUCP>
Date: 4 Mar 90 10:04:28 GMT
References: <100765@looking.on.ca> <534@sixhub.UUCP> <333@comcon.UUCP> <1990Mar3.042312.12638@wolves.uucp>
Organization: Computer Connection, Anchorage Alaska
Lines: 25

In article <1990Mar3.042312.12638@wolves.uucp>, ggw@wolves.uucp (Gregory G. Woodbury) writes:
> In article <103509@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes:
> >The concept of
> >	type foo | tiny-dabe >abe.arc
> >is silly.  It is not meant to be a filter in this manner.  You would use
> >	tiny-dabe <foo >abe.arc
> 
> 	Uhm, Brad.  These two forms are equivalent.

Gregory, the original comment came from me. UUnfortunately, in the
MS-DOS corral, the 2 forms are anything _but_ equivalent.

MS-DOS is single-threaded, so to do the first form, DOS captures the
output of type, writes it to a temporary file (invisible to the user,
because it disappears before you regain control), then executes
tiny-dabe, feeding this temp file to it as stdin. Thus 2 passes and some
extra disk space are used. 

The second form acts in exactly the same way, without the intervening
temporary files. DOS != UNIX... (I wish it did!)
-- 
_R_o_y _M_. _S_i_l_v_e_r_n_a_i_l  | UUCP: uunet!comcon!roy  |  "Every race must arrive at this
#include <opinions.h>;#define opinions MINE  |   point in its history"
SnailMail: P.O. Box 210856, Anchorage,       |   ........Mr. Slippery
Alaska, 99521-0856, U.S.A., Earth, etc.      |  <Ono-Sendai: the right choice!>